DNA mismatch repair (MMR) proteins recognize and excise polymerase errors such as single base misincorporation and insertion/deletion loops during DNA replication. MMR factors also play an important role in meiotic recombination. Mutations in genes encoding MMR and meiotic recombination machinery, have been implicated in a variety of diseases. Understanding the pathways in which these proteins function biochemically will provide mechanistic models that can be used for drug discovery and the development of diagnostic tools. Mlh1-Mlh3 complex is a vital component in a pathway where crossovers form between homologous chromosomes during meiosis. This complex also interacts with Msh2-Msh3 to repair frameshift mutations misincorporated by DNA polymerase. The biochemical steps carried out by Mlh1-Mlh3 in either process are unknown and will be studied here in Saccharomyces cerevisiae, baker's yeast. The first step of meiotic recombination is the formation of double strand breaks, a portion of which are converted into double Holliday junctions (dHJs), which are resolved into crossovers (COs). In S. cerevisiae, genetic evidence suggests dHJ resolution is carried out by Msh4-Msh5, Sgs1 helicase, Exo1 XPG nuclease, and putative Mlh1-Mlh3 endonuclease activitiy. How Mlh1-Mlh3 functions as an endonuclease and how this activity can be used to resolve dHJs is unknown and is the major focus of this project. Our lab's initial characterization of purified yeast Mlh1-Mlh3 complex shows that it is a metal-dependent endonuclease and can nick supercoiled and open circle plasmid DNA. I observed stimulation by Msh2-Msh3, but not ATP or RFC/PCNA. In Aim 1, I will expand on experiments I performed since arriving in the Alani lab to identify the mechanism by which Msh2-Msh3 is stimulating Mlh1-Mlh3. I will also determine if Mlh1-Mlh3 can be strand-directed by the same mechanism as Mlh1-Pms1, the major endonuclease in MMR. Such an activity is a requirement for resolving dHJs into COs. I will also further characterize Mlh1-Mlh3's in vitro substrate requirements. In Aim 2, I will test a proposed dHJ resolution complex involving purified Msh4-Msh5, Sgs1, Exo1, and Mlh1-Mlh3 on model substrates based on the specifications determined in Aim 1. I will also use photo-crosslinking to map protein-protein and protein- DNA contacts as a dHJ resolution complex is assembled. Together these data will allow me to construct a detailed model of how dHJs are resolved into COs in vivo.